A Post on Epistemology
by Elliot Temple

Originally posted on Dec 29, 2002, on the Taking
Children Seriously list:

Imagine someone wrote:

"If many people independently come to hold the same opinion about someone or something and that opinion does not violate basic rules of logic or science, then it probably is correct."

I agree this rule has some aspects of a good rule:
- it's pretty simple
- it's easy to remember
- it doesn't have any unexplained complications or convolutions
- it helps people figure out interesting things

Sadly, one aspect rather ruins the rest - it's not actually true.

I shall explain indirectly. There is a sphere of knowledge called "epistemology". It's the study of the nature of knowledge. And there's something called "our prevailing theories of epistemology" which are the
best ideas about epistemology that humanity has.

These prevailing theories tell us things like:
- knowledge isn't manifest or self-evident
- fallibility (we cannot have entirely certain knowledge)
- authority is invalid (there is no source of knowledge that we should  deem correct in a dispute just because of who it is)
- mass-fallibility (if one person might be wrong about X, and two people might be wrong about X, then a million people might be wrong about X)
- probable knowledge (knowledge with an X% chance of being true) doesn't exist

They also tell us things about how knowledge is justified, like:
- knowledge is not justified from unquestionable foundations
- knowledge isn't justified through being ever so sure
- knowledge isn't justified by prayer or faith
- knowledge isn't justified by authority
- knowledge doesn't come from experience (observations/measurements)
- experience doesn't grant authority
- ideas based on false foundations can still be true
- ideas based on true foundations can still be false
- no justifcation can make a theory entirely certain
- no justification can grant a theory a probability of truth
- good explanations are favoured (and in the quest for new knowledge, so are bold ones that try to say more and are more easily shown false)

Epistemology also helps us judge which explanations are good:
- ones that don't contradict our measurements/observations
- ones without unexplained complications (this one is very important)
- simpler ones
- ones with powerful arguments behind them
- ones that solve problems that exist
- ones that actually do explain the thing in question
- ones that explain more

Also, if two explanations contradict each other, we know at least one is false. So explanations that contradict fundamental aspects of our worldview are either false *or* most of our theories are false. Whether one really wants to start arguing with our most basic of theories is important to consider when creating explanations. For example, one might avoid postulating invisible pink leprachauns because it contradicts the
whole of established physics and to argue it one would have to show  all of physics wrong.

(Read more of Elliot Temple's writing at Fallible Ideas )